[Chrysalis]

I just finished a discussion on irc about ntl been a record slow, here is some figures.

.uk 100mbit 198kB/sec
.nl 10mbit 48kB/sec
.de 100mbit 56kB/sec
adslguide speedtest
Results from broadband speed test recorded on Saturday, 03 December 2005, 16:18.

Downstream 1823 Kbps (227.9 KB/sec) 1968 Kbps (inc. overheads)
Upstream 476 Kbps (59.5 KB/sec) 514 Kbps (inc. overheads)

ntl speed test 7.91mbps

clearly with the differences between ntl speedtest and the other figures there is some kind of external capacity issue with ntl and its a big issue the figures are shocking. All the servers except adslguide I can personally vouch for not been maxed out and almost idle as well as other isp's having no issues with them. I did a trace from the uk box to my ip and I am seeing spikes after enta hands over to ntl over linx supporting that their linx port is saturated.

2 84-45-252-145.no-dns-yet.enta.net (84.45.252.145) 197.386 ms 0.933 ms 0.511 ms
3 telehouse-east.bdr.enta.net (84.45.193.193) 0.369 ms 0.348 ms 0.338 ms
4 linx-gw1.router.ntli.net (195.66.224.22) 0.540 ms 0.428 ms 0.452 ms
5 nth-bb-b-so-600-0.inet.ntl.com (62.253.184.1) 4.940 ms 19.777 ms 5.983 ms
6 lee-bb-a-so-600-0.inet.ntl.com (62.253.185.102) 8.108 ms 24.854 ms 7.703 ms
7 leic-t2core-a-so-000-0.inet.ntl.com (62.253.188.109) 10.681 ms 17.153 ms 9.957 ms
8 leic-t2cam1-a-ge-wan32.inet.ntl.com (82.3.33.9) 32.935 ms 10.230 ms 41.198 ms
9 leic-leic-ubr-8-ge01.inet.ntl.com (82.3.35.150) 10.098 ms 10.184 ms 39.594 ms

note the ping spikes hop 5 onwards. For arguments sake people will tell me icmp is low priority so here is 2 sets of pings, first lot on hop 4 and 2nd lot on hop 5.

64 bytes from 195.66.224.22: icmp_seq=0 ttl=252 time=0.836 ms
64 bytes from 195.66.224.22: icmp_seq=1 ttl=252 time=0.615 ms
64 bytes from 195.66.224.22: icmp_seq=2 ttl=252 time=0.726 ms
64 bytes from 195.66.224.22: icmp_seq=3 ttl=252 time=0.601 ms
64 bytes from 195.66.224.22: icmp_seq=4 ttl=252 time=0.677 ms
64 bytes from 195.66.224.22: icmp_seq=5 ttl=252 time=0.705 ms
64 bytes from 195.66.224.22: icmp_seq=6 ttl=252 time=0.985 ms
64 bytes from 195.66.224.22: icmp_seq=7 ttl=252 time=0.629 ms
64 bytes from 195.66.224.22: icmp_seq=8 ttl=252 time=0.841 ms
64 bytes from 195.66.224.22: icmp_seq=9 ttl=252 time=0.719 ms
64 bytes from 195.66.224.22: icmp_seq=10 ttl=252 time=0.617 ms

64 bytes from 62.253.184.1: icmp_seq=0 ttl=249 time=20.410 ms
64 bytes from 62.253.184.1: icmp_seq=1 ttl=249 time=7.186 ms
64 bytes from 62.253.184.1: icmp_seq=2 ttl=249 time=4.731 ms
64 bytes from 62.253.184.1: icmp_seq=3 ttl=249 time=25.295 ms
64 bytes from 62.253.184.1: icmp_seq=4 ttl=249 time=4.712 ms
64 bytes from 62.253.184.1: icmp_seq=5 ttl=249 time=4.862 ms
64 bytes from 62.253.184.1: icmp_seq=6 ttl=249 time=34.514 ms
64 bytes from 62.253.184.1: icmp_seq=7 ttl=249 time=35.037 ms
64 bytes from 62.253.184.1: icmp_seq=8 ttl=249 time=44.538 ms
64 bytes from 62.253.184.1: icmp_seq=9 ttl=249 time=14.679 ms
64 bytes from 62.253.184.1: icmp_seq=10 ttl=249 time=4.979 ms

This is frustrating since although I expected increased contention speeds of say 3-5mbit but this is quite ridicolous, I am getting performance in some cases on par with a 512kbit connection, people who come on claiming they downloading at 10meg I would be interested in the url they are using and if its of a proxy server, my main interest is non http traffic.